Protein Structure and Dynamics by NMR in Solution
113
description. In addition, the residue 61 requires a chemical exchange term to fulfill statistical tests even when the fully anisotropic tensor is employed. A possible
explanation of the disagreement of the data from helix II, with respect to the rest
of the helical data, is that it has a slightly different mean orientation to that determined in the structure calculation. The homologous sub domain of loop L2 has
been shown to be the last motif to fold in the cytochrome c from mitochondria
(Bai et al. 1995), and is absent from certain families of bacterial cytochromes c
which have otherwise retained very similar secondary and tertiary structural fold
without affecting the stability of the protein (Blackledge et aI. 1996). The loop
may therefore have a certain independence from the rest of the protein fold
which allows a different relative orientation of the helix II. Another possible
explanation for the comparatively low quality fit for this helix would be the presence of low amplitude slow motions of the subdomain L2, hinged at either residues (41 and 69) or residues (45 and 65), although it is difficult to investigate this
type of motion without resorting to unjustifiably complex models.
5
General Conclusions
We have shown that many very different experimental parameters can be
obtained from NMR experiments which furnish valuable information both on the
molecular structure as well as on the dynamical properties of a protein. The multitude of experimental parameters available clearly represents an advantage of
this technique compared for example to radio crystallography. In fact, very exciting applications ofNMR do not only focus on a pure structure determination but
aim to investigate features more related to protein function, as for example
ligand-binding or protein-protein interactions. Molecular interactions can be
monitored by measuring easily obtainable experimental parameters, as for example the chemical shift or heteronuclear relaxation. The possibility of using isotopic filters, by labelling for example only one of the interacting molecules, still
increases the performance of such experiments. In addition, the information
obtained can usually be assigned to individual residues, offering thus the advantage of a very high resolution. A very interesting approach relying on these features of NMR is now routinely used in the industrial research to screen, select
and optimize protein ligands (Shuker et aI., 1996). So hopefully, NMR will find an
increased entrance into pharmacological and biological research where it should
prove to be a very versatile technique for the study of protein properties related
to biological function.
6
Acknowledgements
This work has been supported by the Centre National de la Recherche Scientifique, the Commissariat a l'Energie Atomique and Molecular Simulations Inc. We
thank Drs GJ Arlaud and JF Hernandez as well as M Cusanovich and M Caffrey
for kindly providing us with Clr-EGF and Rhodobacter capsulatus cytochrome C2>
respectively, and for the fruitful collaboration that was established between our
Précédent

- 121/371

Suivant